Recent efforts in analyzing the HMP (Human Microbiome Project) data have uncovered numerous microbiota small molecules that are significantly different between healthy and diseased humans. However, it is challenging to unravel their contribution to diseases, largely due to the lack of an efficient system to manipulate their levels in vivo. Based on our recent finding that we can toggle specific microbiome molecule in vivo using bacterial genetics, we start with the HMP data to develop an integrative and generalizable approach to predict their metabolic genes, modulate their levels in the host using CRISPR-Cas9-based bacterial genetics, and investigate their effects on host intestinal immunity. Our approach will efficiently delineate the microbial gene-molecule- host phenotype connection and will lay the groundwork for generating a gut microbial community with defined and editable molecular outputs for disease prevention and cure. There are three convergent motivations for this work: First, many microbiota molecules are different between healthy and diseased humans, and we need functional studies that causally connect these molecules with a disease. Using the HMP data, we will design an integrative and generalizable approach to modulate the level of a single microbiome molecule in the host while leaving others barely changed. This approach will boost a systematic identification of disease-causing microbiota metabolites, our findings will also promote new therapeutic strategies by targeting these previously unknown microbial metabolic pathways. Second, microbiome metabolites are one of the most intriguing phenotypes conferred by the gut microbiota to the host, and interrogating their interaction at the molecular level will uncover new biology. Our approach will examine the effects of microbiota metabolites on host biology in a rational and systematic way. This approach can be applied to engineer microbiome molecular flow at different body sites where host and microbe crosstalk to decipher their downstream biological effects. Our findings would also open the door to controlling one of the most concrete contributions that gut bacteria make to host biology. Third, an engineered microbial community with a defined and programmable molecular output has therapeutic potential. The microbiome is part of our ?pan-genome?, whose metabolic outputs are more tractable by genetic manipulation. Our proposed approach will facilitate the characterization of the microbiome metabolites and their connection with host biology, which form the basis for a synthetic community with a defined and programmable molecular output to prevent and cure diseases.